Details
The feature uses a custom generative AI model built on Stable Diffusion and ControlNet, trained by Linearity's machine learning team for approximately 240 hours on an A6000 Ada GPU. It takes a raster image as input, isolates the foreground subject, and generates a new background that accounts for lighting, shadows, and reflections. Users can choose from preset theme categories (such as Minimalist, Nature, or Cityscapes) or write their own text prompt of up to 400 characters.
Have evidence about Linearity (Vectornator)'s AI practices? Submit a report.
Report a Sighting →